9. Working with Evolutionary Algorithms

In this chapter we discuss the practical aspects of using EAs. Working with EAs often means comparing different versions experimentally, and we pro- vide guidelines for doing this, including the issues of algorithm performance measures, statistics, and benchmark test suites. The example application (Sect. 9.4) is also adjusted to the special topics here; it illustrates the application of different experimental practices, rather than EA design.

Contents:
9.1 What Do You Want an EA to Do? ……………………147
9.2 Performance Measures ……………………………..150
9.2.1 Different Performance Measures …………………151
9.2.2 Peak Versus Average Performance ……………….155
9.3 Test Problems for Experimental Comparisons ……………158
9.3.1 Using Predefined Problem Instances ……………..158
9.3.2 Using Problem Instance Generators………………160
9.3.3 Using Real-World Problems…………………….160
9.4 Example Applications………………………………161
9.4.1 Bad Practice………………………………..161
9.4.2 Better Practice………………………………162

Suggested Reading

Exercises

The on-line accompaniment to the book Introduction to Evolutionary Computing